Social status and relationships affect language production. Consider the converse: can socio-cultural information be inferred from linguistic evidence? Specific phenomena naturally vary by language, so an intermediate variable, language use, abstracts over diverse phenomena such as formal/informal pronouns (Russian vy), indirect/direct phrasing (English), or related features (e.g. Navajo fourth person, Faltz 1998 p. 32). Typical scales for socio-cultural analysis use labels or natural numbers (1, 2, 3, etc.) as in teineigo, sonkeigo, and kenjogo for Japanese keigo (Wetzel 2004, p. 4). Labels do not translate well, but neither do natural numbers: The bottom level "1" in each language may describe dissimilar situations; the middle level in one language may be 3 but in another 2. What scale might work cross-linguistically? We introduce a scale with "zero" as a neutral level, developed for closeness and deference in Tamil, Korean, and Russian. We also examine French, Hindi, and English examples.